Preliminary phylogeny of the thrips parasitoids of Turkey based on...

279

Research Article

Turk J Zool34 (2010) 279-289 © TÜBİTAKdoi:10.3906/zoo-0906-29

Preliminary phylogeny of the thrips parasitoids of Turkeybased on some morphological scales and 28S D2 rDNA, with

description of a new species

Oğuzhan DOĞANLAR1,*, Mikdat DOĞANLAR2, Anne FRARY3

1Ağrı İbrahim Çeçen University, Faculty of Science and Arts, Department of Biology, Ağrı - TURKEY2Mustafa Kemal University, Faculty of Agriculture, Department of Plant Protection, Antakya, Hatay - TURKEY

3İzmir Institute of Technology, Department of Molecular Biology and Genetics, Urla, İzmir - TURKEY

Received: 12.06.2009

Abstract: Species of the Ceranisus thrips-attacking genus are difficult to distinguish morphologically. The phylogeneticrelationships within the Ceranisus species were explored using nucleotide sequences of the 28S D2 expansion region ofthe rDNA gene. Bayesian, maximum likelihood, and parsimony inference methods were employed to construct thephylogenetic relationships. Principal component analysis on the Turkish species of Ceranisus, namely antalyacus, menes,bozovaensis, hirsutus, planitianus (a new record for Turkey), pacuvius, and a new species, provided supporting evidence.All known data concerning hosts and geographical distribution are presented. A new species, C. onuri O. Doganlar, sp.n.,was described from Turkey.

Key words: Thrips, parasitoids, Ceranisus, phylogeny, 28S rDNA, Turkey

Türkiye thrips parazitoitlerinin bazı morfolojik ölçümler ve 28S D2 rDNAözelliklerine dayanarak oluşturulan ön filogenisi ile yeni bir türün tanısı

Özet: Thripsler üzerinde doğal düşman olan, Ceranisus cinsinin morfolojik olarak teşhis edilmeleri oldukça zordur. Busebeple çalışmada bu böceklerin rDNA’sına ait genişlemiş 28S D2 bölgesinin gen dizisi kullanılarak, Ceranisus türleriiçin bir filogenetik ilişki ortaya konmuştur. Filogenetik ilişki oluşturulurken Bayesian, Maximum likelihood ve Parsimonymetodları kullanılmıştır. Morfolojik ölçümler kullanılarak Ceranisus, viz., antalyacus, menes, bozovaensis, hirsutus,planitianus (Türkiye için yeni kayıt), pacuvius ve yeni bir tür üzerinde temel bileşenler analizi yapılmış ve sonuçlar bu yenitürün farklı tür olduğunu desteklemiştir. Türler üzerinde coğrafik yayılış, ilişkide olduğu konak bitkiler ile ilgili bilinentüm bilgiler verilmiştir. Türkiye’den yeni bir tür C. onuri O. Doganlar, sp.n. tanımlanmıştır.

Anahtar sözcükler: Thrips, parazitoit, Ceranisus, filogeni, 28S rDNA, Türkiye

* E-mail: [email protected]

IntroductionThrips (Thysanoptera: Thripidae) are very small

insects distributed worldwide with approximately5000 species identified so far. Approximately a fewhundred thrips species are serious pests of crops dueto the injury they cause by feeding on flowers or seedcapsules, which results in bruising, deformity, andreduced yield of the crop (Lewis, 1997). Due to theirsmall size, special behavior, and high rate of bioticpotential, thrips are difficult to control by classicalcontrol methods. Only the species in 2 families,Eulophidae: Entedontinae and Trichogrammatidae,of parasitoid Hymenoptera are known to parasitizetheir larvae and eggs, respectively. All of the thripsparasitoids belonging to Entedontinae are found in 5genera (Boucek, 1976, 1988; Schauff, 1991; Triapitsynand Headrick, 1995; Doğanlar, 2003; Noyes, 2007).

The genus Ceranisus Walker (Hymenoptera:Eulophidae), the most important monophyletic groupin Entedontinae, are larval parasitoids of thrips, some ofwhich are used for biological control of thrips pests(Schauff, 1991; Loomans and van Lenteren, 1995;Triapitsyn and Headrick, 1995; Loomans and Pakozdi,1996; Loomans, 2003). The Nearctic species of Ceranisuswere reviewed by Triapitsyn and Headrick (1995), andTriapitsyn (2005) gave a world taxonomic revision ofCeranisus and listed their known host associations.Noyes (2007) listed 16 species of Ceranisus in worldfauna. The identification key of European species ofCeranisus was first provided by Graham (1963), whoalso described Ceranisus lepidotus Graham(Hymenoptera: Eulophidae) from England. Erdös(1966) described Ceranisus planitianus Erdös(Hymenoptera: Eulophidae) from Hungary. The firstthrips parasitoids species in Turkey was found byDoğanlar (2003), who described Urfacus bozovaensisDoğanlar (Hymenoptera: Eulophidae), which wasrecently synonymized by Doğanlar and Triapitsyn(2007) as Ceranisus bozovaensis (Doğanlar). Cameronet al. (2004) described a new species, C. antalyacus S.Triapitsyn, 2004, from the western Mediterraneanregion of Turkey. Doğanlar and Triapitsyn (2007)published a taxonomic revision of Turkish species ofCeranisus and provided identification keys to both sexesof the European and Turkish species of Ceranisus. Theyalso described a new species, Ceranisus hirsutusDoganlar and S. Triapitsyn (Hymenoptera: Eulophidae),and redescribed C. bozovaensis, including the female.

Morphology alone may not be enough todistinguish all species clearly, and molecular dataprovide an additional useful tool for identification ofspecies. Molecular data have proven to be verypowerful for resolving difficult problems (Chen et al.,2004). Using DNA sequence data, Campbell et al.(2000) intended to resolve the relationships ofChalcidoidea. Sequence data for 28S D2 rDNA wereused to construct a phylogeny of the familyEulophidae (Gauthier et al., 2000). However, there arevery few studies using molecular data of rDNA forCeranisus species; only the Kenyan isolate of C. meneswas previously sequenced by Gauthier et al. (2000).

This paper is the first phylogenetic study ofCeranisus based on DNA sequence data from thesecond expansion segment (D2) of the 28S ribosomalsubunit. The goals of the study were to perform amodern phylogenetic classification of Turkish speciesof the genus, and to distinguish and identify speciesbased on their morphology, includingmorphometrics, by performing principal componentanalysis.

Materials and methodsTaxon samplingOur study focused on the species of Ceranisus that

parasitize larvae of some thrips pests in Turkey. Insectsampling and collection details are listed in Table 1.During 2005-2007, a total of 174 Ceranisus specimenswere collected from thrips larvae on their hosts insouthern and southeastern Turkey. The specimenswere killed in ethanol and subsequently stored in afreezer (–20 °C), and some were slide-mounted inCanada balsam. The examined specimens weredeposited in the Insect Museum of the PlantProtection Department (ICMKU), Faculty ofAgriculture, Mustafa Kemal University, Antakya,Hatay, Turkey.

Principal component analyses of morphologicalmeasurements

Specimens were examined for a range of discreteand continuous morphological variables. Themorphological character measurements of theantennae, wings, and ovipositors of females were usedto determine the factors of the principal components.

Preliminary phylogeny of the thrips parasitoids of Turkey based on some morphological scales and 28S D2 rDNA, with description ofa new species

280

All data were found to be normally distributed, andprincipal component analyses were undertaken usingthe PCA option of XLSTAT software (Addinsoft Ltd,Paris, France).

DNA amplification and sequencingGenomic DNA was extracted from single

specimens, using the DNeasy Tissue Kit (QIAGEN),by freezing and heating methods (QIAGEN, Leusden,the Netherlands). Sequence fragments displaying anincreasing degree of variability were analyzed in theconserved D2 expansion of the 28S nuclear gene.Standard 50 μL PCR reactions were performed using0.2 U Taq DNA polymerase (Fermentas), 5 μL of 10×Taq buffer with KCl (Fermentas), 3 μL of 25mMMgCl2 (Fermentas), 1 μL of 10× dNTPs (Fermentas),1 μL of each primer, 2 μL of template DNA, and 36.8μL of dH2O (Sigma). Primer sequences for the 28SrDNA D1F (ACC CGC TGA ATT TAA GCA TAT)and D2R (TTG GTC CGT GTT TCA AGA CGG)were from Harry et al. (1998) and Campbell et al.(1993), respectively. PCR conditions for C.bozovaensis, C. hirsutus, and C. antalyacus were 30cycles of 94 °C denaturation (30 s), 55 °C annealing(30 s), and 72 °C elongation (90 s), with an initial 94°C denaturation (3 min) and a final 72 °C extension(30 min). For C. menes, C. planitianus, and the newspecies, the conditions were 30 cycles of 94 °Cdenaturation (30 s), 62 °C annealing (30 s), and 72 °Celongation (60 s), with an initial 94 °C denaturation(3 min) and a final 72 °C extension (10 min). AfterDNA amplification, 5 μL of product with 1 μL ofloading dye (Fermentas) was loaded onto 1% agarosegel to check the DNA amplification. The remainingDNA was loaded onto a 1.5% agarose gel withethidium bromide, separated by electrophoresis at 140V for 1 h, and visualized under UV light. Theamplified product was purified using a QIAquick GelExtraction Kit (QIAGEN). DNA fragments were runon an ABI 3100 Avant Genetic Analyzer (AppliedBiosystems, Foster City, CA, USA).

Sequence alignment and gene analysisSequences were aligned with the program

ClustalW (Thompson et al., 1994) in MEGA 3.1(Kumar et al., 2004) using the default setting (Opengap penalty = 15, extend gap penalty = 6.66). Thealignments were checked manually.

Equally weighted maximum parsimony (MP) andmaximum likelihood (ML) analyses were performedusing PAUP (4.0 beta version) (Swofford, 2000). ForMP analysis, a heuristic search procedure was usedwith the following default settings: 10 replicates ofrandom taxon addition, tree-bisection-reconnectionbranch swapping, multiple trees retained, no steepestdescent, and accelerated transformation. Gaps weretreated as missing data. For ML analysis, theappropriate substitution model of DNA evolution thatbest fit (GTR-gamma model) the dataset wasdetermined by Akaike’s information criterion withModel Test 3.06 (Posada and Crandall, 1998) asfollows: 28S rDNA, base frequencies = (A: 0.1468; C:0.3033; G: 0.3321; T: 0.2178), rate matrix = (1.9391;2.1793; 1.2682; 0.3533; 3.9780; 1.0000), and gammadistribution shape parameter = 0.2806. Bootstrapanalysis with 1000 replicates was calculated as ameasure of support for individual clades for the MPand ML trees. A Bayesian analysis was conducted oneach dataset using Metropolis-coupled Markov chainsMonte Carlo (MCMCMC) implemented in theMrBayes software (Huelsenbeck and Ronquist, 2001).Each Bayesian analysis consisted of 4 chains, randomstarting trees, a uniform prior distribution ofparameters, and the GTR+I+_ model of nucleotidesubstitution. The chains were run for 2 milliongenerations, and trees were sampled every 100generations. The best tree was determined by visualexamination of the log-likelihood plots and the burn-in trees were discarded. As the MP and ML analysesusually gave multiple trees, we reduced the set of treesto one consensus tree. The sequences of Galeopsomyiasp. (AJ274458) and Cirrospilus sp. (Eulophidae:Eulophinae) (AJ274438; Gauthier et al., 2000) wereused as out-groups for these analyses.

ResultsExamination of morphological charactersThree factors were retained for the principle

component analyses (PCA), and their cumulativepercentages were calculated as 54.69%, 69.51%, and82.87%, respectively. The first 2 factors, accountingfor 69.51% of the cumulative variance, were plottedagainst each other (Figure 1). The eigenvalues,variance and cumulative variance, analysis result,eigenvectors, and factor loading values of the 3

O. DOĞANLAR, M. DOĞANLAR, A. FRARY

281

282


principle components are shown in Table 2. SixCeranisus species were clearly separated in thecoordinate systems. C. antalyacus, C. menes, C. onurisp.n., and C. hirsutus were located in 4 different polesof the scatter plot. C. planitianus was intermediatebetween C. antalyacus and C. menes, while C.bozovaensis was located between C. onuri sp.n., C.pacuvius, and C. hirsutus. While the differentspecimens of most Ceranisus species were closelyclustered, the C. bozovaensis samples were morewidely distributed in the coordinate system. The scapeand first flagellar segment length to width ratio andwing characteristics were of primary importance forthe first principle components. Ovipositor length andantennal characters were determined by PCA as thesecond principle components (Table 2) foridentification of the Ceranisus species of Turkey.

28S D2 rDNA sequence data of CeranisusThe size of the amplified 28S D2 expansion region

of the rDNA gene fragment ranged from 415 to 576bp. The base composition of the sequence had astrong bias toward cytosine and guanine, whichconstituted approximately 62.3% of the total. Thealignment of the sequenced fragment resulted in 464characters, including gaps. Of these, 345 characterswere constant, 59 characters were variable and

parsimony-uninformative, and 60 characters wereparsimony-informative. The alignment was relativelystraightforward and did not require insertion gaps.The base composition of the 28S D2 region was asfollows: A, 0.143; C, 0.301; G, 0.322; T, 0.233. Theslight G bias evident in the 28S sequences was notedin Chalcidoidea by Gillespie et al. (2005) and can beattributed to guanine’s ability to base pair with bothcytosine and uracil in RNA molecules (Gutell et al.,1994).

Maximum parsimony analysis of the amplified 28SD2 expansion region of the rDNA gene fragmentproduced 156 equally parsimonious trees [consistencyindex (CI) = 0.9038, retention index (RI) = 0.8421]and bootstrap (1000 replicates). The 50% majorityrule strict consensus tree is given in Figure 2. Theheuristic search yielded one single tree with a score of-Ln likelihood = 1395.15050 (Figure 2b). There aresome differences between the consensus treesgenerated under maximum parsimony and maximumlikelihood, but in general, the topologies of thesereconstructions are similar. The 50% majority ruleconsensus tree inferred from the Bayesian analysis isvery similar to the results of the parsimony andlikelihood analyses (Figure 2c). There is one conflictbetween the Bayesian trees and the parsimony trees;they merely differ in the separation of C. hirsutus. TheBayesian posterior probabilities offered a slightlyhigher nodal support in comparison with thebootstrap analysis.

Parsimony, maximum likelihood, and Bayesiananalyses produced similar topologies (Figure 2). Themain differences were in the position of the groupincluding C. menes, C. planitianus, C. pacuvius, andC. onuri sp.n., and the other group including C.antalyacus, C. hirsutus, and C. bozovaensis. In theanalysis, high bootstrap values (91%-100%) weredetermined between C. menes and C. planitianus, andthey were located on different branches of the trees.In addition, 2 species were clearly separated fromother Ceranisus species, with 100% bootstrap andBayesian posterior probabilities scores. Ceranisuspacuvius and C. onuri sp.n. occurred in one group andwere located on the same branch in all trees, but theywere clearly separated by bootstrap values (100%) and20 bp differences in maximum parsimony analyses.The parsimony tree showed 29 bp differences between

5

4

3

2

1

0

-1

-2

-3

-4

-5

hirsutus

-- ax

is F2

(15.1

9%)--

bozovaensis

antalyacus

planitianus

pacuvius

onuri

menes

>

-- axis F1 (55.37%)-->5 4 3 2 1 0 1 2 3 4 5- - - - -

Figure 1. Scatter plot for Ceranisus individuals distributed on acoordinate system of the first 2 factors resulting fromprincipal component analysis.

Tabl

e 1.

The

num

ber o

f Cer

anisu

sspe

cies

, sam

plin

g lo

calit

ies,

altit

ude,

colle

ctio

n da

te, a

ssoc

iate

d pl

ants,

mos

t com

mon

thrip

s spe

cies

on

host

plan

ts,th

e EM

BL G

enBa

nkac

cess

ion

num

ber o

f Tur

key

thrip

s par

asito

ids,

and

thei

r DN

A se

quen

ces.

Sp

ecie

sL

oca

lity

Coord

inate

Alt

.D

ate

Sp

eci.

Th

rip

sH

ost

pla

nt

Acc

ess.

no.

(o,ı

an

d N

:E)

(m)

Cera

nisu

s ant

alya

cus

Hat

ayAl

tınöz

ü-Ya

nıkp

ınar

36 0

1: 3

6 13

720

10.0

4.07

2♂, 2♀

Taen

ioth

rips i

ncon

sequ

ens

Arbu

tusi

ncon

sequ

ens

EU55

7271

Yayl

adağ

ı-

10.0

4.07

5♀

Yayl

adağ

ı-

20.0

4.07

4♀

Cera

nisu

s onu

riN

iğde

Mad

en-K

arag

öl37

25:

34

3324

2517

.06.

0611♀

Thrip

s mer

idio

nalis

As

phod

eline

dam

asce

na

EU55

7272

Bolk

ar M

ount

ains

37 2

6: 3

4 35

2182

17.0

6.06

23♀

Cera

nisu

s boz

ovae

nsis

Şanl

ıurfa

Bozo

va37

22:

38

2968

507

.04.

071♂

,3♀

Thrip

s tab

aci,

Whe

at, l

entil

, chi

ckpe

aEU

5572

73

07.0

4.07

3♂,5♀

;H

aplo

thrip

s trit

ici,

23.0

4.06

2♀H

aplo

thrip

s reu

teri

Kang

örm

ez36

26:

38

1144

504

.05.

063♀

Adıy

aman

Göl

başı

37 3

7: 3

7 30

857

25.0

7.06

1♀

Cera

nisu

s hirs

utus

Şanl

ıurfa

Kang

örm

ez36

26:

38

1144

506

.04.

072♂

,5♀

Unkn

own

Whe

at, l

entil

, lee

kEU

5572

74

Adıy

aman

Fıra

t riv

er

37 2

7: 3

8 15

402

25.0

3.07

2♀pl

ants,

and

som

e gra

sses

Cera

nisu

s pac

uviu

sK.

mar

aşCe

nter

14.0

5.03

2♀Un

know

nM

edica

go sa

tiva a

nd

EU55

7275

Şanl

ıurfa

Bire

cik-A

rat M

ou.

37 0

3: 3

8 08

800

06.0

5.05

2♂,9♀

lentil

plan

ts

Bozo

va37

22:

38

2968

507

.04.

073♂

,5♀

;

Aslan

lı vi

llage

37 4

6: 3

9 26

937

04.0

5.06

3♀

Cera

nisu

s pla

nitia

nus

Gaz

iant

epAk

yoku

ş37

10:

36

5811

0025

.03.

072 ♀

Unkn

own

Leek

plan

ts an

d EU

5572

76

Hat

ayPa

yas-

Sinc

an Y

ayla

sı36

43:

36

1449

005

.05.

072♀

som

e gra

sses

Cera

nisu

s men

esG

azia

ntep

Islah

iye

36 5

6: 3

6 34

518

16.0

3.07

1♀Th

rips t

abac

i, W

heat

, len

til, c

hick

pea,

AJ23

4445

Arab

an-Y

.mül

k37

27:

37

2692

123

.05.

071♀

Hap

loth

rips t

ritici

leek

plan

ts, an

d so

me

Adıy

aman

Göl

başı

37 3

7: 3

7 30

857

24.0

7.06

1♀gr

asse

s

Hat

ayAt

çana

36 1

4: 3

6 22

9630

.04.

0714♀

Tekn

epın

ar36

07:

35

5626

030

.04.

072♀

283


62

100 100

100

100

C. antalyacus

C. bozovaensis

C. hirsutus

C. pacuvius

C. onuri

C. planitianus

C. menes

Galeopsomyia sp.

Cirriuspilus sp.

a

100

100 97

100

C. antalyacus

C. bozovaensis

C. hirsutus

C. pacuvius

C. onuri

C. planitianus

C. menes

Galeopsomyia sp.

Cirriuspilus sp.

b

52

56

100

100

91

100

C. antalyacus

C. bozovaensis

C. hirsutus

C. pacuvius

C. onuri

C. planitianus

C. menes

Galeopsomyia sp.

Cirriuspilus sp.

c

Figure 2. a) Bayesian tree, using MCMCMC method and GTR+I+_ model; b)Bootstrap 50% majority rule consensus tree inferred by maximumlikelihood analyses of the 28S D2 rDNA gene (GTR-gamma model,-Ln likelihood = 1395.15050); c) The strict consensus mostparsimonious tree, inferred by 28S D2 rDNA sequences. Numbersabove lines represent bootstrap values from 1000 replicates on allparsimony-informative characters, and only bootstrap values >50%are shown. Tree length is 156, CI is 0.9038, HI is 0.1630, RI is0.8421.

284


Table 2. The eigenvalues, variance, statistics, eigenvectors, and factor loading values for the first 3 factorsresulting from principle component analysis.

F1 F2 F3

Eigenvalue 6.645 1.823 1.522

% variance 55.373 15.192 12.684

Cumulative % 55.373 70.565 83.248

Chi-square (observed value) 693.300

Chi-square (critical value) 85.965

DF 66

One-tailed P-value <0.0001

Alpha 0.05

Eigenvectors Factor loadings

F1 F2 F3 F1 F2 F3

Scape 0.121 0.338 0.640 0.311 0.456 0.790

Pedicel 0.145 –0.310 0.398 0.375 –0.419 0.491

F1 0.271 –0.299 0.164 0.699 –0.403 0.202

F2 0.357 –0.065 0.192 0.922 –0.087 0.237

Clava 0.308 –0.332 0.209 0.795 –0.448 0.257

C1 0.303 0.416 –0.097 0.782 0.561 –0.119

C2 0.361 0.184 –0.001 0.930 0.249 –0.001

Specula 0.314 –0.009 –0.410 0.808 –0.012 –0.506

Forewing 0.287 –0.170 0.008 0.740 –0.229 0.009

Longest marginal cilia 0.332 –0.238 –0.253 0.855 –0.321 –0.312

Hind wing 0.334 0.115 –0.274 0.860 0.156 –0.338

Ovipositor 0.208 0.529 0.089 0.537 0.715 0.110

285


populations of the C. onuri sp.n. and C. planitianusgroup and group 1, and the likelihood analysesdemonstrated a 12-base difference between the samepopulations of Ceranisus species. In likelihoodanalyses, a single 28S substitution supporting therelationship between C. hirsutus, C. bozovaensis, andC. antalyacus was homoplasious, with very lowbootstrap support and 1 or 5 bp differences.

DiscussionC. menes and C. planitianus are morphologically

different from other Ceranisus species of Turkey asthey have almost smooth heads and mesosoma and

bear distinct semioval bare areas at the posteriormargin behind the base of the marginal vein. C. menesis a solitary endoparasitoid of thrips larvae and isfound on a wide range of host plants in differentbiotopes around the world (Loomans, 1991; Goodwinand Steiner, 1996; Triapitsyn, 2005; Doğanlar andTriapitsyn, 2007). In Turkey, it was collected fromweeds, vegetable crops, barley, lentils, chickpeas, leeks,and some grasses in approximately all samplinglocalities of the Mediterranean region. C. planitianusis distributed in Canada, the USA, Israel, Turkey (newrecord), and Europe (Loomans and Van Lenteren,1995; Triapitsyn, 2005; Triapitsyn and Morse, 2005),but was rarely found in collection areas in Turkey. In

286


our study, C. planitianus was found on Raphanus sp.with white flowers in only 2 localities with 4 femalespecimens. C. menes differs from C. planitianus inhaving the malar groove whole and straight, andmetasoma yellow or light brown (in C. planitianus, themalar groove is Y-shaped and the metasoma is darkbrown or black). Our PCA and molecular analysessupport the morphological data. Thus, this differencewas clearly shown in PCA results; the points based onthe morphological characters of specimens of each ofthese species were closely grouped and the speciesclusters were located in different areas of the scatterplots. Additionally, the divergence between C. menesand C. planitianus was strongly supported by MP, ML,and Bayesian analyses with high bootstrap values(91%, 97%, and 100% respectively) and by concordantbase changes in the 28S D2 rDNA region. Although C.onuri sp.n. specimens were collected from differenthabitats, climatic conditions, and geographic regionsof Turkey compared to C. pacuvius (Table 1), geneticanalyses show good bootstrap support for the high(94%) association between C. onuri sp.n. and C.pacuvius. An adjusted distance matrix shows that C.onuri sp.n. and C. pacuvius differ in 5 characters (basepairs), indicating that these 2 species are related.These results were corroborated by PCA analysesbased on selected morphological characters of C.menes and C. pacuvius. These 2 species were separatedfrom other Ceranisus species under the effect of thefirst principle components, which were length of thespicula of C2, long marginal setae, and forewinglength to width ratio, and were located in the positivepole of the F1 axis. According to the morphologicalcharacters of this species, C. onuri sp.n. differs from C.lepidotus in having an almost smooth body, while C.lepidotus has a reticulated body. These differenceswere clearly shown in the PCA scatter plot and theMP, ML, and Bayesian analyses consensus trees. Theother 3 species were placed in the same group by thegenetic analyses, but all examined species have quitedifferent morphological characters. C. hirsutus has 3segmented clavae of female antenna and 17 setae onthe mesosoma, C. antalyacus has a very differentfemale antennal scape with C1 longer than C2(Cameron et al., 2004; Doğanlar and Triapitsyn,2007), and C. bozovaensis has a notably expandedmarginal vein of the male forewing. C. hirsutus and C.bozovaensis have similar habitat choices, e.g., weeds,

wheat, lentils, and alfalfa plants, whereas C. antalyacuswas obtained from different habitat in Turkey. Onlythis species was collected together with Taeniothripsinconsequens, feeding on flowers of pear and Arbutusandrachne. Those differences were shown only in theMP analyses of the Ceranisus species. C. antalyacus,C. bozovaensis, and C. hirsutus were separated by MPanalyses with 52 and 56 bootstrap scores and 8, 2, and5 base pair differences, respectively. However, thesedifferences were not supported by the maximumlikelihood or Bayesian analyses (Figure 2).

In conclusion, molecular data for the amplified28S D2 expansion region of the rDNA gene fragmentproduced an informative phylogeny of the genusCeranisus. The molecular data are consistent with themorphological data for supporting the differences ofC. menes, C. planitianus, C. onuri sp.n., and C.pacuvius. In contrast, the phylogenetic relationshipsamong members of C. antalyacus, C. hirsutus, and C.bozovaensis are not completely supported by themorphological data and results of PCA analysis andfurther analyses are needed. However, the 28S D2rDNA phylogeny obtained by maximum parsimony,maximum likelihood, and Bayesian analyses generallyagreed with the relationships based on previousmorphological criteria.

Ceranisus onuri O. Doğanlar, sp.n.(Figures 3–7.)Types. Holotype ♀ (on slide), labeled: “TURKEY,

Niğde, Maden, Karagöl, 37°25′N, 34°33′E, 2425 m,05.05.2005, O. Doganlar. Mounted at ICMKU by thefirst author 2007 in Canada balsam.” Paratypes, 11♀♀, same data as the holotype; 23 ♀♀, Niğde, BolkarMountains, 37°26′N, 34°35′E, 2182 m, 17.06.2006, M.Doğanlar: stored at –20 °C in 96% alcohol, ICMKU.

DescriptionFemale: Body length 0.7-1.16 mm (holotype 1.06

mm). Head and mesosoma brown; antenna, legs,venation, and metasoma light brown; wings hyaline.

Head. Vertexal suture broadly V-shaped (Figure 3).Antenna (Figure 4) sparsely setaceous with scapeslender, about 6.5× as long as wide, pedicel notablyshorter than scape, 2.6× longer than wide, F1 almostcylindrical and without sensillae, shorter and narrowerthan F2, 0.43× as long as and 0.4× as wide as F2; F2 with

3

4

6

7

5

Figures 3-7. Ceranisus onuri sp.n. female. 3) Head, 4) Antenna, 5) Mesoscutum andscutellum, 6) Forewing, 7) Gaster and genitalia.

287


1 sensillum; clava 2-segmented, including spicula 3.02×as long as wide, C1 slightly shorter than C2.

Mesosoma (Figure 5) 0.69× as long as metasoma;pronotum and mesoscutum in 1/5 apical area withlight engraved reticulation; each side lobe ofmesoscutum with one seta; midlobe of mesoscutum

with 2 pairs and scutellum with 1 pair of setae.Forewing (Figure 6) 3× as long as wide; uniformlycovered with numerous microtrichia; longestmarginal cilia about one-fourth maximal width offorewing; submarginal vein with 2 long macrochaetaeand 3 hypochaetae opposite to basal macrochaetae;

Boucek, Z. 1976. Changes in the classification of some AfricanChalcidoidea (Hymenoptera). J. Ent. Soc. Sth. Afr. 39: 9-31.

Boucek, Z. 1988. Australasian Chalcidoidea (Hymenoptera). ABiosystematic Revision of Genera of Fourteen Families, with aReclassification of Species. CAB International Institute ofEntomology, Cambrian News Ltd., Aberystwyth, Wales.

Cameron, E.A., Teulon, D.A.J., Triapitsyn, S.V. and Tunç, I. 2004. Thediscovery of a new species of Ceranisus from southwesternTurkey. BioControl 49: 373-383.

Campbell, B.C., Steffen-Campbell, J.D. and Werren, J.H. 1993.Phylogeny of the Nasonia species complex (Hymenoptera:Pteromalidae) inferred from an internal transcribed spacer(ITS2) and 28S rDNA sequences. Insect Mol. Biol. 2: 225-237.

Campbell, B.C., Heraty, J., Rasplus, J.Y., Chan, K., Steffen-Campbell,J. and Babcock, C. 2000. Molecular systematics of theChalcidoidea using 28S-D2 rDNA. In: The Hymenoptera:Evolution, Biodiversity and Biological Control. (ed. A.D. Austinand M. Dowton), CSIRO Publishing, Melbourne, pp. 59-73.

Chen, Y., Xiao, H., Fu, J. and Huang, D. 2004. A molecular phylogenyof eurytomid wasps inferred from DNA sequence data of 28S,18S, 16S, and COI genes. Mol. Phylogenet. Evol. 31: 300-307.

Doğanlar, M. 2003. A new genus and a new species of Entedontinae(Hymenoptera: Eulophidae) from southeastern Anatolia,Turkey. Turk. J. Zool. 27: 181-185.

Doğanlar, M. and Triapitsyn, S.V. 2007. Review of Ceranisus(Hymenoptera: Eulophidae) of Turkey, with description of anew species. Eur. J. Entomol. 104: 105-110.

Erdös, J. 1966. Nonnullae Eulophidae novae Hungaricae(Hymenoptera, Chalcidoidea). Annales Historico-NaturalesMusei Nationalis Hungarici (Pars Zoologica) 58: 395-420.

Gauthier, N., LaSalle, J., Quicke, D.L.J. and Godfray, H.C.J. 2000.Phylogeny of Eulophidae (Hymenoptera: Chalcidoidea) with areclassification of Eulophinae and the recognition thatElasmidae are derived eulophids. Syst. Entomol. 25: 521-539.

Gillespie, J.J., Yoder, M.J. and Wharton, R.A. 2005. Predictedsecondary structures for 28S and 18S rRNA fromIchneumonoidea (Insecta: Hymenoptera: Apocrita): impact onsequence alignment and phylogeny estimation. J. Mol. Evol. 61:114-137.

Goodwin, S. and Steiner, M.Y. 1996. Survey of Australian nativenatural enemies for control of thrips. Bulletin IOBC/WPRS 19:47-50.

Graham, M.W.R. de V. 1963. Additions and corrections to the Britishlist of Eulophidae (Hym., Chalcidoidea), with descriptions ofsome new species. Trans. Soc. Br. Entomol. 15: 167-275.

Gutell, R.R., Larsen, N. and Woese, C.R. 1994. Lessons from anevolving rRNA: 16S and 23S rRNA structures from acomparative perspective. Microbiol. Rev. 58: 10-26.

References

288


postmarginal vein 1.98× as long as stigmal vein,marginal vein and parastigma 6.04× as long as stigmalvein. Hind wing about 6.87× as long as wide; bladeuniformly setaceous.

Metasoma (Figure 7) with petiole almost as longas wide. Ovipositor occupying about two-thirdslength of gaster, slightly exserted; ovipositor tometatibia length ratio 1:1.2.

Relative measurements, as length or length/width:Antenna scape: 13/2, pedicel: 6.5/2.5, F1: 3/2, F2: 7/5,clava: 9.6+1.3/3.5, C1: 4, C2: 5.5, spicula: 1.3, forewing50/17, longest marginal cilia 4.2, hind wing 44/6.4,ovipositor: 22.3.

Diagnosis:By following the key of Doğanlar and Triapitsyn

(2007) for the European and Turkish species ofCeranisus, the new species, C. onuri, runs to thefemale of C. pacuvius, but it differs from C. pacuviusin having a vertexal suture broadly V-shaped (Figure3) (in C. pacuvius, a vertexal suture broadly C-shaped); clava, including spicula 3.02× as long as wide

(in C. pacuvius, clava including spicula 2.2× as longas wide); and forewing (Figure 7) 3× as long as wide(in C. pacuvius, forewing 2.5× as long as wide).

Male: unknown.

Biology: larval parasitoid of Thrips meridionalis(Priesner) (Thysanoptera: Thripidae) living onAsphodeline damascena (Boiss.) at 2500 m elevationin the Taurus (Bolkar) Mountains.

Acknowledgements

We wish to thank Assoc. Prof. Dr. İlhan Üremis,Mustafa Kemal University, Faculty of Agriculture,Department of Plant Protection, Hatay, foridentification of our plant species. We also wish tothank Dr. Laurence Mound, CSIRO Entomology,Canberra, ACT 2601, Australia, for checking ourthrips specimens. This work was supported by theScientific and Technological Research Council ofTurkey – TÜBİTAK (TOVAG-106 0 244).


289

Harry, M., Solignac, M. and Lachaise, D. 1998. Molecular evidencefor parallel evolution of adaptive syndromes in fig-breedingLissocephala (Drosophilidae). Mol. Phylogenet. 9: 542-551.

Huelsenbeck, J.P. and Ronquist, F. 2001. MrBayes: Bayesian inferenceof phylogeny. Bioinformatics 17: 754-755.

Kumar, S., Tamura, K. and Nei, M. 2004. MEGA3: an integratedsoftware for molecular evolutionary genetics analysis andsequence alignment. Brief Bioinform. 5: 150-163.

Lewis, T. 1997. Thrips as Crop Pests (CD package). CABInternational, Oxford, New York.

Loomans, A.J.M. 1991. Collection and first evaluation ofhymenopterous parasites of thrips as biological control agents ofFrankliniella occidentalis. Bulletin IOBC/WPRS 14: 73-82.

Loomans, A.J.M. 2003. Parasitoids as biological control agents ofthrips pests, PhD dissertation, Wageningen University,Wageningen, the Netherlands, 200 pp.

Loomans, A.J.M. and Pakozdi, A. 1996. Differential acceptance ofCeranisus menes (Walker) and Ceranisus americensis (Girault)attacking thrips hosts, Frankliniella occidentalis (Pergande) andThrips tabaci (Lind.). Folia Entomol. Hung. (suppl.) 57: 83-90.

Loomans, A.J.M. and Van Lenteren, J.C. 1995. Biological control ofthrips pests: a review on thrips parasitoids. In: BiologicalControl of Thrips Pests (eds. A.J.M. Loomans, J.C. van Lenteren,M.G. Tommasini, S. Maini and J. Riudavets), WageningenAgricultural University Papers, 95-1. Wageningen, VeenmanDrukkers, pp. 89-193 & 195-201 (Appendix).

Noyes, J. 2007. Universal Chalcidoidea Data Base. http://www.nhm.ac.uk/jdsml/research-curation/projects/ chalcidoids.

Posada, D. and Crandall, K.A. 1998. MODELTEST: testing the modelof DNA substitution. Bioinformatics 14: 817-818.

Schauff, M.E. 1991. The Holarctic genera of Entedoninae(Hymenoptera: Eulophidae). Contrib. Am. Entomol. Inst. 26:1-109.

Swofford, D.L. 2000. PAUP: Phylogenetic Analysis Using Parsimony(*And Other Methods). Sinauer, Sunderlands, MA, USA.

Thompson, J.D., Higgins, D.G. and Gibson, T.J. 1994. CLUSTAL W:improving the sensitivity of progressive multiple sequencealignment through sequence weighting, position-specific gappenalties and weight matrix choice. Nucleic Acids Res. 22: 4673-4680.

Triapitsyn, S.V. 2005. Revision of Ceranisus and the related thrips-attacking entedonine genera (Hymenoptera: Eulophidae:Entedoninae) of the world. Afr. Invertebr. 46: 261-315.

Triapitsyn, S.V. and Headrick, D.H. 1995. A review of the Nearcticspecies of the thrips-attacking genus Ceranisus Walker(Hymenoptera: Eulophidae). Trans. Am. Entomol. Soc. (Phila.).121: 227-248.

Triapitsyn, S.V. and Morse, J.G. 2005. A review of the species ofCeranisus Walker (Hymenoptera: Eulophidae) in the NewWorld. Trans. Am. Entomol. Soc. (Phila.) 131(1+2): 69-86.

Preliminary phylogeny of the thrips parasitoids of Turkey based on...

Documents

Transcript of Preliminary phylogeny of the thrips parasitoids of Turkey based on...